Structural Impacts of Research
This chapter analyzes the structural implications of publicly funded research and technology development for U.S. agriculture. It first examines academic analyses. Then, the chapter addresses the structural implications of specific types of innovations, including mechanical, chemical, biologic, and managerial. Third, the chapter describes the structural impacts associated with combinations of innovations in three case studies that examine the Green Revolution, the introduction of the tomato harvester, and new developments in animal husbandry. Finally, this chapter discusses the structural implications of the process of setting priorities in agricultural research, including the criteria used and methods of obtaining stakeholder input.
RESEARCH AND THE STRUCTURE OF AGRICULTURE
Very little empirical evidence exists on the relationship between public sector agricultural research and structural change. Two studies examined the correlation between the extent or intensity of public research in the United States and the rate of increase in farm size and related structural characteristics. The first study, by Busch and colleagues (1984), examined the U.S. Censuses of Agriculture and other U.S. Department of Agriculture (USDA) data by state for 1915–1973. Indicators of the public research effort (including research expenditures and the number of research personnel) were related statistically to
several indicators of farm size and to the concentration of agricultural production. A simultaneous-equation model was employed, using the state as the unit of analysis. The study provides strong statistical evidence that publicly financed agricultural research and development (R&D) is correlated with increases in average farm size, the number of very large farms (1,000+ acres), and large farms as a percentage of all farms (when controlling for a range of variables, such as farm mortgage debt, government payments to farmers, or value of marketed farm output). The largest effect is seen in the increase in the relationship between R&D expenditures and the percentage of large farms.
Another study, by Huffman and Evenson (2001), used a similar database and reached comparable findings. Huffman and Evenson used data from 1950 to 1982 from the Censuses of Agriculture, USDA, and related state-level sources. The authors used a six-equation econometric model and a large number of control variables to disaggregate the factors that lead to structural changes in farming. The objective was to estimate proportional differences over time in farm-structure-dependent variables attributable to three sets of variables: public R&D and education, private R&D and market forces, and farm commodity program payments. The indicators of structural change included crop and livestock specialization, an index of average farm size (essentially a normalized indicator of the average value of services obtained from physical capital and farmland), and amount of part-time farming. Huffman and Evenson reported that public research and education have been at least as important as private research and development and market forces for changing livestock specialization, farm size, and farmers’ off-farm work participation over the study period. The strength of the relationship between public research and farm growth increased over the last third of the study period (from roughly the early 1970s to the early 1980s). Private R&D and market forces have been relatively more important than public research and education for changing crop specialization. Changes in farm commodity programs had little relationship to farm structure over the study period.
Although the Busch et al. (1984) and Huffman and Evenson (2001) studies represent different disciplines (rural sociology and agricultural economics, respectively), they have largely consistent results. In the aggregate, they associated the extent or intensity of the public agricultural R&D effort with an increased scale of agricultural production. However, these studies analyzed a limited set of variables (e.g., number, size, specialization, and farmers’ off-farm work participation), whereas the study committee used a much broader definition of structure that includes a wide range of variables. Thus, it is difficult to make a general statement about the overall relationship between public-sector R&D and structural change based on these results. Furthermore, the Huffman and Evenson study demonstrates that the relationship of public R&D to different variables is mixed depending on the structural variable tested. They present evidence to support that public R&D is a major factor for some structural variables but not
for others. It is thus difficult to assess the magnitude of the relationship between public R&D and structural change relative to that of other drivers of structural change.
INNOVATION AND THE STRUCTURE OF AGRICULTURE
The literature suggests that innovations that result from research vary in their influence on the structure of agriculture (Sunding and Zilberman, 2001; Thirtle and Ruttan, 1987). We can distinguish among mechanical, biologic, chemical, and managerial innovations, which also can be divided into those that increase yield or reduce costs of farming. The cost-reducing innovations can be subdivided into those that are labor saving and those that are capital saving. Related categories include innovations that augment human capital (automated management strategies) or that preserve natural resources. Modern irrigation technology, for example, can improve land quality and conserve water (Caswell, 1991). With the rise in consumerism, the importance of product-based innovations has grown, and there is much effort to improve the quality of food products. A related category of innovation is improved postharvest performance of agricultural systems, for example, that extend the shelf life of fruits and vegetables or that streamline shipping and handling. The environmental movement has raised the value of protecting environmental quality and of reducing the damage caused by agricultural activities. An additional category of environmental innovations detects damage and promotes better monitoring of farming and ecosystem performance (Millock et al., in press). Innovations in satellite imaging are expected to improve environmental decision making in agriculture (NRC, 1997b). This variety of innovations could have implications for the structure of agriculture. This section discusses innovations that influence horizontal consolidation, vertical integration, and regional distribution.
The concept of scale neutrality is used as a criterion for assessing the structural impact of a particular innovation. Scale neutrality is the ease with which a particular technology can be adopted, its divisibility into small enough units to be adopted, and its potential to benefit large and small producers alike in terms of results or relative profit. Scale neutrality is often confused with, but is distinct from, divisibility (see Chapter 3). Many divisible technologies, such as improved seed, which can be obtained in small or large quantities at the same unit price, are not necessarily scale neutral if they cannot be applied to a small-scale context.
Mechanical innovations typically are applied to farm machinery. Tweeten (1989) argued that mechanical innovations might have the greatest influence on horizontal consolidation, especially with regard to the increase of farm size and the reduction in the number of farms. Mechanical innovations contribute to horizontal consolidation for two reasons: First, they tend to reduce the requirement for labor—a main input provided by the farmer. Second, the capital cost per acre declines as the size of the farm increases. Capital-cost advantages favor large farmers for additional reasons, as well. Owners of larger farms often have easier access than do small operators to the capital they need to invest in equipment.
Mechanical innovations were important in our transformation from an agrarian to an urban society. Tractors and harvesting equipment saved labor, and they contributed to the increase in the size of farm operations (Kislev and Peterson, 1996). The trend continues, as new machinery dictates an increase in the size a farm must be to achieve economic viability—particularly in field-crop production.
There is a perception that large commercial farms are capital intensive and that smaller farm operations are more labor intensive. Although this is true in most cases, some of the large-farm operations in the country—Dudda Brothers in Florida and some major fruit and vegetable growers in California, for example—are very labor intensive. U.S. agriculture relies heavily on immigrant labor, particularly for harvesting, because the wage is too low to be appealing to many native-born American workers. Dependence on immigrant labor creates the problem of legal and illegal immigration. Automation of harvesting can actually lead to a reduction in manual, low-skill jobs and create higher paying jobs that attract nonimmigrant workers and, as is the case with the California lettuce industry, involvement with organized labor (Martin, 1985; Martin and Perloff, 1997).
Labor-saving machinery, such as the cotton harvester, has had other important social implications. Automation and mechanical innovations have been important for the viability of part-time farmers (Bessant, 2000; Huffman and Evenson, 2001; Kislev and Peterson, 1996). Parker and Zilberman (1996) have found that part-time farmers who own small orchards are among the first to adopt the use of computerized irrigation. Caswell et al. (1984) demonstrated that these types of operations are also among the first to adopt drip irrigation.
The operation of agricultural machinery, especially tractors, has been identified as among the most hazardous occupational activities in the United States and around the world (Forastieri, 1999). Some industries, such as the mining industry, tend to encourage the substitution of capital for labor to maximize worker safety. So innovations that increase safety and improve the
well-being of farm workers can make the industry more attractive to potential employees.
Mechanical innovations are important in fostering environmentally sustainable forms of agriculture. For example, equipment is required for the transition from conventional tillage to reduced tillage systems. Some mechanical innovations, including those that use computers, can be important for monitoring the environmental impact of agricultural systems. Mechanical innovations also are being put to use for waste management.
Despite the importance of the mechanical sector in agriculture, the public research contribution in this area has not been substantial. Most mechanical innovations, many of them in farm machinery, have been introduced by the private sector (Feder et al., 1985). As we demonstrate in Chapter 4, only a small amount of public expenditure is devoted specifically to mechanical innovations for agriculture, although work supported by other public entities, including the military, the National Aeronautics and Space Administration, and the Department of Energy, could have spillover effects in agriculture.
Chemical innovations, such as the development of herbicides that tend to replace labor, are likely to favor large farms, although that would mainly be the result of volume discounts for procurement. The volume used per acre would be the same for large and small farms. Application costs per acre likely decline with size, but that phenomenon is a consequence of mechanical application (Feder et al., 1985; Thirtle and Ruttan, 1987).
Examination of the structural effects of chemical innovations in agriculture shows mixed effects: Most chemical solutions are divisible, easy to apply, and usable even by the smallest farms. Mid-size or small family farms with field crops (corn, cotton) often are large enough to afford aerial spraying, and many chemicals are applied by certified applicators who charge on a size-neutral, per-unit basis. Some larger farmers receive volume discounts or own their equipment, so they save resources in application, but these have not been documented to be major advantages. The simplicity and low labor intensity of chemical treatments can benefit mid-size family farms operated by older farmers. On the other hand, the adjustment and education associated with pesticide regulation and the transition to new pesticide application techniques all require human capital and effort, which could favor the more dynamic commercial farms and lead to older farmers’ exit from farming (Green, 1995; McWilliams and Zilberman, 1996; Putler and Zilberman, 1988). Finally, the use of chemicals, particularly herbicides, allows fewer operators to manage more acres.
Much of the basic chemical and engineering research that led to the use of pesticides, fertilizers, and chemicals occurred outside colleges of agriculture and agricultural experiment stations. Most recent pesticide research has been
performed in the private sector. Annual sales of chemical pesticides in the United States were $12 billion in 1997, of which about 7 to 13 percent is estimated to have been spent on private research and development. Development, testing, and registration of a chemical can take 8 to 12 years and can cost over $50 million for each pesticide (NRC, 2000b). Land grant scientists and Agricultural Research Service (ARS) researchers do not emphasize basic research in chemistry. The public sector invests in pest management research but seldom in research on chemical toxicity and effectiveness. Public-sector work complements that of the private sector, and in some cases, the chemical industry may pay researchers or specialists to test new products.
Innovations in Biology
Innovations in biology, especially new seed varieties, increase general yield and can actually increase yield per acre. Their use requirements do not vary with the size of the farm. On those grounds, they have more neutral structural impacts than do mechanical or chemical innovations. On the other hand, some innovations in biology, for example, the new tomato varieties introduced to complement the tomato harvester, have significant structural effects in that larger farms might have an economic advantage in adopting new innovations because of fixed costs associated with education, capital requirement, and other factors (Feder et al., 1985). The literature on adoption suggests that when new crop variety properties differ significantly from those of traditional varieties, adoption can require drastic changes in the production system (Mann, 1978). For example, adoption of Green Revolution high-yielding varieties1 often requires fertilizers and irrigation to be profitable and effective. Green Revolution varieties have been associated with high fixed costs for education and adjustment, which in the short term confer a competitive advantage to more affluent, better educated growers (Thirtle and Ruttan, 1987). Introduction of new varieties will have a smaller structural effect when the new varieties are closely related to old ones. Introduction of more drastically different varieties can have differential effects according to farm size.
New applications of biology to agriculture have resulted in seeds that obviate the use of chemical pesticides (Bt cotton2) or that augment their use (Roundup Ready varieties3). Innovations in biotechnology could permit gradual
modification of existing varieties and, thus, might not require significant adjustments or changes in the structure of the system. Bt corn and Roundup-resistant soybeans, for example, were adopted at high rates because adoption did not require significant adaptation of the production system. Pray (1993) suggested that the genetically modified varieties have higher rates of adoption and are accepted by smaller farmers because they are simple and convenient to use. Convenience, low cost, timesaving, and simplicity have been cited as reasons for the high adoption rates for genetically modified varieties in the United States (Carpenter and Gianessi, 1999; Fulton and Keyowski, 1999; Hubbel et al., 2000). However, more stringent environmental regulations associated with biotechnology would likely reduce adoption and produce fixed-cost and knowledge requirements that could deter some farmers. In addition, as chemicals, pharmaceuticals, and other exotic materials are increasingly produced using biotechnologic tools, vertical integration and contract farming are likely to result as the private sector buys the rights to and develops marketing strategies for these innovations (Zilberman et al., 1999).
Innovations in biology also could significantly affect the structure of agriculture because of differential benefits across regions. Olmstead and Rhode (1993) have demonstrated that innovations in biology have enabled the crop production system to adapt to different ecologic zones and have contributed to the growth of agricultural productivity in the United States. New heat- or cold-tolerant traits can permit the expansion of production of certain crops (Caswell et al., 1984). For example, drought-tolerant Hass avocadoes and cherry tomato varieties have permitted production of those crops in arid regions of California. Similarly, frost tolerance traits have expanded strawberry production. We could see a trend toward consolidation and vertical coordination in agriculture if new farm operations in these locations are larger or are part of agribusiness organizations, as was the case with the expansion of tolerant varieties of specialty fruits to major growers in the West. However, introduction of new varieties that benefit small local growers in some regions could have the opposite effect.
Increased product differentiation in agriculture and the introduction of new varieties, each with unique features and attributes, can contribute to increased value added and profitability of agriculture. Innovations in biology could be an important source of increased diversity among crop varieties. Increased differentiation among varieties will contribute to increased differentiation among agricultural products and could enable farmers to have a wider choice of agricultural inputs and ways to respond to variations in weather, soil, and pest conditions. However, introducing new differentiated varieties to the market is not easy and could require vertical coordination. Many farmers will be unwilling to adopt new varieties that result in a distinct and differentiated product unless there is a buyer for this product. A private company that owns the right to genetic materials is likely to contract with farmers to grow the product and then buy it from them to sell it down the marketing chain. Calgene, for example,
contracted with growers for the cultivation of Flavr Savr tomatoes4. New differentiated varieties can increase value added to agriculture and are also associated with increased contract farming.
Public research historically has emphasized developing and improving genetic materials (Busch et al., 1995). Because capturing the benefits from new seeds was difficult for the private sector, genetic selection and variety improvements fell mainly to public-sector research. The role of the private sector has expanded over time, however, beginning with the introduction of hybrids in the 1930s. The introduction of plant breeders’ protection legislation in the 1970s (U.S. Congress, 1970) and the Supreme Court decision in 1980 (Diamond v. Chakrabarty, 1980) that allowed patenting of life forms gave private companies the legal tools to protect their investments in developing genetic materials (Wright, 1998). Indeed, we have seen an emergence of seed industries that coexist with the public sector in producing genetic materials.
The public-sector research and development effort is linked closely to the Consultative Group of International Agriculture Research (CGIAR) centers, which house seed banks and other genetic materials and conduct exchange programs among breeders in different countries. Alston et al. (1995) documented the significant economic benefits the United States has obtained from investments in new varieties and strains developed in other countries. However, access to genetic diversity could become limited and more expensive with the privatization of genetic materials.
Changes in prices, weather, technology, institutions, and personnel can influence resource allocation and profits in the agricultural sector. Effective managerial practices consider all of those factors. Research on many management practices is needed to prevent faulty resource allocations, reduce the public and private costs of errors that do occur, and increase efficient use of resources. Jensen (1977) reported that managerial research focuses on assisting decision makers to use resources efficiently; helping policy makers determine the consequences of alternative policies; studying the economic effects of technologic and institutional changes on agricultural production and resource use; and studying individual farm, area, and regional adjustments in resource use.
Managerial innovations help farmers run their operations. One recent example is precision farming, which uses Geographic Information System/Global Positioning System (GIS/GPS) technology to ascertain field characteristics and minimize the use of irrigation or fertilizer, for example. Because of the high initial investment in equipment, this technology has been
affordable only for large farms. Savings in input use may offset the initial outlay, but there is no consensus on this. Some studies argue that precision farming results in increased yields, reduced input use, and reduced environmental damage from excessive chemical use (Kitchen et al., 1996; Koo and Williams, 1996; NRC, 1997b; Sawyer, 1994; Watkins et al., 1998), but other studies demonstrate mixed results of precision farming on profitability (Carr et al., 1991; Swinton and Lowenberg-Deboer, 1998). Profitability studies may not be conclusive until spatial econometrics, whole-farm analysis, and management information system analysis are included in economic analyses (Olson, 1995). The effect could be mitigated if equipment suppliers are willing to provide custom service or rental arrangements (NRC, 1997b).
Vertically integrated structures, particularly in animal production, provide another example of managerial innovation. The relatively new broiler production industry, for example, developed from contractual arrangements between the grain industry and poultry farmers in which safe markets were ensured for the broiler producers in exchange for their guarantees of feed purchase from the grain industry (McBride, 1997). Hog farming also exhibits a high degree of vertical integration (Martinez, 1999). Vertical integration demonstrates mixed structural impacts: It can benefit small producers by reducing overall business risk, controlling costs, gaining and improving market position, and facilitating access to information and financial resources necessary to develop new crops. However, contract farming creates its own risk, despite reducing others. This risk is associated with failure to produce to contract standards, loss of independence, and weak bargaining power in negotiating contracts. Therefore, farmers who are unwilling to enter contractual agreements may be forced out of business (Rehber, 1998).
Socioeconomic data from government sources can be used to enhance private and public managerial decisionmaking. USDA’s Economic Research Service (ERS) and National Agricultural Statistics Service (NASS) regularly compile and publish data on prices, quantities supplied, and quantities consumed, for example. USDA’s Agricultural Marketing Service (AMS) also provides more time-sensitive information on a daily basis through its Market News Program. Those data can be used by farmers in making decisions, especially about production. Although the data are publicly available and their use would seem to be scale neutral, the value of the information for various farmers depends on their ability to access and use the data. The factors that affect decisions to adopt technology discussed in more detail in Chapter 3.
Advances in computer and information technology have been dramatic over the past 15 years and have contributed significantly to reducing the costs of management throughout the global economy, including the agricultural sector (World Bank Group, 2000). Putler and Zilberman (1988) analyzed survey data from Tulare County, California to demonstrate that size of the farming operation, education level, age level, and ownership of a farm-related nonfarming business significantly influence the probability of computer ownership. The type of
application software owned is influenced primarily by the type of farm products produced, the size of the farming operation, ownership of a farm-related business, and the education level of the farmer. Other studies have found similar relationships between education and age level of users and the adoption of computers and the number of software applications used (Batte et al., 1990; McWilliams and Zilberman, 1996).
It is difficult to judge the scale neutrality of managerial innovations. Some (precision agriculture) appear to benefit larger farms. Others (vertical integration) have led to large structural changes in animal production and show mixed effects on the production abilities of smaller farms. Other innovations (price information) appear scale neutral but depend on farmers’ ability to adopt them. Thus, managerial innovations can cause significant structural changes, but not necessarily to the exclusive benefit of larger farms.
Structural impacts associated with several important applications of combinations of innovations can be described.
The impacts of public research on farm size and structure were addressed in the debate over the social consequences of the Green Revolution (see Lipton and Longhurst, 1989, for comprehensive review).5 The Green Revolution debate was initiated about South Asia (particularly India and Pakistan), an area in which there were pronounced inequities in agricultural resources and rural political power. The use of short-stature “miracle wheats” in South Asia from the mid-1960s through the early 1970s resulted in extraordinary yield increases, typically a two-fold or larger increase in output per hectare over a short period (2–5 years). Dramatic increases in yield and the potential for applying the technology to famine relief in South Asia led to wheat breeder Norman Borlaug’s winning the Nobel Peace Prize in 1970—perhaps the all-time high-water mark of global public research in agriculture. A concurrent trend toward increased loss of land
and increases in farm size and concentration was observed in South Asia. The critics of the South Asian Green Revolution argued that landlessness and increased wheat output were causally related and that similar effects of the Green Revolution were beginning to occur in the other major Green Revolution crops and regions6. Similar effects were observed in Asian rice and in Latin American maize production (Pearse, 1980). Proponents of the Green Revolution argue that on the whole, benefits of reduced food prices and increased food security outweighed the adverse structural impacts (Ruttan, 2000). Heated debate over the structural implications of the Green Revolution continues to this day.
At virtually the same time that concerns were first being raised about possible structural consequences of the Green Revolution (see the history in Lipton and Longhurst, 1989), Jim Hightower’s Hard Tomatoes, Hard Times chronicled similar issues with regard to U.S. public research7. Hard Tomatoes, Hard Times was not a formal academic study, but it suggested that the role of the University of California in developing the mechanical tomato harvester, which would be substituted for farm worker labor, was an example of how public research could have structural effects. Most land grant administrators and many land grant scientists criticized Hightower’s exposé of the land grant research and extension system, generally claiming that land-grant-developed technologies are usually scale neutral, and thus unbiased.
While for many observers, the Hightower book first raised the issue of the structural impacts of public research on tomato harvest mechanization, the still-classic study on the topic by Schmitz and Seckler (1970) had been published several years earlier. In addition to the Schmitz and Seckler study, there has been a large empirical literature on public research and the scale of agricultural production in the tomato sector (Berardi, 1984; DeJanvry et al., 1980; Friedland and Barton, 1975; Friedland et al., 1981). These studies show that in response to the threatened termination of the Bracero Program8, which provided inexpensive Mexican labor to California tomato producers, University of California agricultural engineers assisted in bringing to market a mechanical harvester that largely mechanized the harvest of processing tomatoes. These studies show that
the development and adoption of the tomato harvester dramatically altered the structure of the processing tomato sector, resulting in declines in the numbers of farms9 and increases in the average scale of production and in the concentration of this sector. By contrast, there was relative stability in the structure of the fresh tomato sector where the tomato harvester was not widely adopted.
There is little doubt that the University of California’s role in the development of the tomato harvester contributed directly and decisively to the increased scale of tomato production in California (particularly in processing tomatoes). It is not clear that the public sector was fully responsible for increasing the scale of agriculture in this case, for several reasons. First, during the 1960s, and even today, land grant, State Agricultural Experiment Station (SAES), or ARS funding of farm mechanization research has been relatively small. Thus, the University of California funding of harvest mechanization equipment research was anomalous by national standards. Second, tomato harvest mechanization had some social benefits, such as reduced consumer prices (Schmitz and Seckler, 1970). Thus, even the most well-researched case studies of the impact of public research on farm scale and structural change do not lend themselves particularly well to answering a basic question that is part of the focal point of this report.
There has been relatively more research on adoption of new technology and methods to produce cultivated crops than there has been concerning animal agriculture. However, the most significant structural changes in agriculture have occurred in the livestock sector. The broiler industry and the swine industry, for example, present a new mode of industrial agriculture characterized by contracting, vertical integration, high concentration of animals, and increasing returns to scale. The dairy sector has experienced significant changes; large dairies in southern California, New Mexico, and Texas (with several hundreds or even thousands of cows that rely on prepared feed) have become an increasingly dominant segment of the market.
Many technologic and institutional innovations that led to increased regionalization, concentration, and vertical coordination in livestock production originated in the private sector. Public-sector research contributed little to technologies such as automated milking machines or to the herringbone-milking parlor, which increased the size of dairy farms. Similarly, automated poultry feeding systems resulted from mechanical innovations developed in the private sector in the United States and abroad.
Public-sector research could have contributed significantly to other areas of industrialized animal production. Data from research on diet, genetics, metabolism, and digestion were used by commercial producers to design feed formulae and develop industrialized animal production systems. Similarly, new information about disease prevention and animal health control facilitated increased animal density. Publicly supported discoveries on the manipulation of light to increase the egg-laying productivity of hens were also important in providing the economic rationale for industrialized egg production. In addition, public research institutions have contributed technologic innovations that increased the industrialization of agriculture. At Iowa State University, boars were selected and tested for their performance in confinement on concrete floors. Leg weakness and hoof shape were modified through breeding to improve their suitability for concrete floors (Hargrove, 1973; Rothschild and Christian, 1988).
Recent debate and scholarly investigation of the effects of public-sector research on the structure of agriculture have arisen surrounding the use of biotechnology, including recombinant bovine somatotropin (rbST)10 and genetically engineered crop varieties. Biotechnology issues include matters of scale neutrality and scale bias and the implications of large-scale research agreements such as that consummated in 1999 between Novartis (now Syngenta) and the University of California, Berkeley11.
Examination of the structural impacts of biotechnology, particularly rbST, has yielded two major findings. First, the extent of the public and private contributions to those technologies is extremely difficult to disaggregate. In the case of rbST, other than the Cohen-Boyer research on cloning genetically engineered molecules in cells (Cohen et al., 1973), most R&D was done commercially by Genentech, Monsanto, Cyanamid/American Home Products, Elanco, and others. Nevertheless, although there was relatively little public funding, SAES and land grant university scientists and research facilities were pivotal in the development of the technology. Public researchers tested rbST on their own herds and allocated scientist effort to its development. Later, Cooperative Extension encouraged its adoption.
A second major finding is that although rbST technology is divisible into small enough units, in principle, to be used on a farm of any scale, the pattern of rbST adoption is highly correlated with herd size. Using a comprehensive data set on rbST adoption in the United States, a 1999 Wisconsin study reported that
more than 70 percent of operators with herds of 200 or more cows used rbST (Buttel et al., 2000; Ostrom and Buttel, 1999). However, only about 4 percent of operators with herds of fewer than 50 cows used rbST. One critical feature of rbST use is that its benefits can be maximized only if high-quality feed is available and animal nutrition is managed accurately. Adoption of rbST is highly correlated with the use of other productivity-augmenting technologies (e.g., total mixed ration [TMR] equipment, which blends all feedstuffs [forage, grain, and supplements] to provide a complete source of nutrients in a ration) found almost exclusively on large dairy farms (Buttel et al., 2000). Thus, somewhat similar to Green Revolution crops, rbST is a highly divisible but scale-biased practice in terms of adoption rates across farms of different scales of production.
STRUCTURAL IMPLICATIONS OF THE RESEARCH PRIORITYSETTING PROCESS
Criteria for Setting Priorities in Agricultural Research
The criteria used for setting priorities in public-sector agricultural research and the assessment of the payoff or benefits in the form of return on investment can have significant structural effects. Many of the criteria and payoff assessments have focused on productivity and efficiency goals, justified on the assumption that in itself, increased productivity will benefit producers and consumers and will feed the expanding global population. We now know that the issue is more complex. Hunger has many causes, including inadequate income to purchase food and distribution systems that are inadequate to transport food from producer to consumer. Achieving a goal of simply producing more food masks other problems related to expanding populations, including increases in disease and the destruction of natural resources vital to a productive and resilient ecosystem. Increased productivity in a market characterized by inelastic demand (where prices respond dramatically to small changes in quantities supplied) significantly reduces income for producers even when output increases, often reducing producer income to unacceptable levels.
The tension between serving diverse constituencies while promoting rapid, and often socially dislocating, increases in productivity is still at issue in public agricultural research well over a century after the public agricultural research and outreach system was established. The goal of publicly funded agricultural research and outreach in the United States as originally articulated in three acts of Congress—the Morrill Act of 1862 (U.S. Congress, 1862), the Hatch Act of 1887 (U.S. Congress, 1887), and the Smith-Lever Act of 1914 (U.S. Congress, 1914)—was to serve rural people as a whole. Together, these acts promoted:
“[T]he liberal and practical education of the industrial classes” (Morrill)
“[A] sound and prosperous agricultural and rural life,” ensuring “agriculture a position in research equal to that of industry, which will aid in maintaining an equitable balance between agriculture and other segments of the economy” (Hatch); and
“[The] diffusing among people…[of] useful and practical information on subjects relating to agriculture and home economics, and [encouraging] the application of the same.” (Smith-Lever)
Congress intended to use public agricultural research and outreach to help farmers and the “mechanic classes” to advance. Thus, social goals and service to a broad constituency were emphasized relative to goals of productivity or efficiency. These goals were an attempt to respond to the generally limited farmer support for research in the late nineteenth century. Farm journalists, chemists, and university administrators supported passage of research legislation (Danbom, 1986). In contrast, American farmers were more interested in seeking relief from their economic troubles (by curbing railroad monopolies, obtaining credit, expanding exports) than they were in supporting research or new technology (Marcus, 1985). In particular, the Hatch Act, which explicitly directed funds to be used for applied research, represented a compromise between the aims of its proponents (to modernize agriculture, “professionalize” the farmer, and harness science as an engine of national development) and the concerns of its opponents (to be practical and of broad benefit to rural people; Marcus, 1985).
It is important to acknowledge those who benefit in the long term from productivity and efficiency research in a commodity industry such as agriculture. Although the short-term benefits of advances in technology that improve productivity and efficiency can accrue to producers and other participants in the production and distribution chain, in a private-sector commodity industry characterized increasingly by global competition, the benefits are eventually captured by consumers, if no monopoly conditions exist. Although the focus of productivity-enhancing R&D might be on producers, the rest of the value chain, including consumers of food products, captures the payoff over the long term. Furthermore, the producer segment that captures the most benefits in a commodity industry is the one with the lowest expenses or the greatest control. If an industry is characterized by decreasing cost (as appears to be the case in agricultural production), then larger scale producers will naturally capture more of the benefits of productivity-focused research. We do not suggest here that productivity-focused research is inappropriate. We propose only that in a commodity industry characterized by decreasing cost and intense competition, the economic forces inherent in the market will in the long run transfer the benefits of productivity- and efficiency-increasing research to other parts of the
value chain. Benefits retained will be those by the lowest cost, largest scale producers.
A second structural implication of productivity-efficiency criteria for funding and evaluating research is the limitation of these criteria to consider broader social goals in assessing the benefits of agricultural research. Those goals might include increasing diversity in agricultural production and distribution systems; reducing environmental and resource degradation; contributing to the long-term sustainability of agricultural production systems; improving the social well-being of producers and rural residents; and reducing financial, economic, political, and environmental risk. Only the public sector addresses these goals; their benefits cannot be captured by the private sector. Increasingly, this broader set of criteria is part of the allocation system for public-sector agricultural R&D expenditures. However, the fundamental distribution issue of who gains, who loses, and how the losses might be mitigated or repaid has never been a criterion for evaluation of public-sector research projects. Similarly, distributional consequences of research are not the focal point of the agenda for evaluating specific advances or innovations in agriculture.
As a larger proportion of the R&D budget moves into the private sector, the total public- and private-sector R&D budget in agriculture will focus increasingly on innovation that ignores the broader set of social and public goals but benefits consumers and efficient, large-scale producers. This will occur to the extent that productivity and efficiency criteria dominate the allocation and assessment process and as public-sector funding declines in relative proportion to private-sector funding of R&D. Private-sector R&D will be evaluated almost exclusively for productivity and efficiency, because that is how value is created and captured, and managers of publicly held companies are evaluated by generation of profits and the market value of stock. It should be acknowledged that, as segments of agriculture move from a commodity orientation to a differentiated-product orientation, lowest cost is no longer the only or dominant determinant of competitive advantage, and innovation that contributes to differentiation (rather than to cost reductions) has the potential to capture value. If producers of all sizes can adopt such an innovation, it can be less size biased than a productivity-efficiency innovation in a decreasing-cost, competitive industry. Whether there will be size advantages to producing differentiated versus commodity products and who will capture the benefits from innovation in differentiated product markets remain unanswered questions. Furthermore, as long as most of the differentiation in the food production and distribution chain occurs beyond the farm gate, the benefits from innovation in differentiated products will be captured by those who produce that differentiation and not by farmers or producers.
A third structural implication of the productivity-efficiency criteria for research funding and evaluation is the inadequacy of narrowly defined criteria to
measure total resource productivity and efficiency. Specifically, the criteria are typically measured and defined in terms of private cost. Because of important differences between private and public cost, including such externalities as long-term resource degradation and the social cost of human resource adjustments, the private-cost-driven, productivity-efficiency criteria will be biased against R&D that might reduce total resource productivity and efficiency but that increase long-term sustainability. Ignoring public cost as part of the criteria for R&D allocations and assessment will bias the allocation process toward innovations that reduce private cost at the expense of those that might reduce both public and private cost. In essence, even if efficiency and productivity are the only social goals to be achieved in public-sector R&D allocations, ignoring externalities and public cost will have important distributional and structural implications.
The committee encourages the public sector to develop broader criteria for evaluating and funding agricultural research that will help producers—particularly those producers outside mainstream agriculture who are unable to compete in commodity markets—obtain and retain market value. Those goals might include increasing diversity in agricultural production and distribution systems; reducing environmental and resource degradation; contributing to the long-term sustainability of agricultural production systems (see Pretty, 1995 for a more detailed discussion of sustainable agriculture); improving the social well-being of producers and rural residents; and reducing financial, economic, political and environmental risk. The committee recognizes that there are limits to the degree to which developing technology for “niches” is sustainable, since increased research and development on a niche product will increase the size of the market, invite entry by other producers, and thereby turn the niche product into a commodity product. The quest for higher value niche production technology and products is thus a perpetual one. Nevertheless, the committee notes that the relative contribution of public research in developing technologies relevant to small-scale farmers, organic farmers, and others outside the commercial mainstream is an important determinant of its support to those constituencies. Structural concerns should be better balanced with all other factors involved in setting the research agenda. Only the public sector can address these goals; their benefits cannot be captured by the private sector.
The goals of public-sector research should continue to be broadened beyond productivity and efficiency. Federal and state research should improve technology and information systems that benefit farmers in diverse production systems and circumstances, including part-time farmers, small-scale farmers, organic farmers, and value-added producers. However, limiting public-sector research to scale-neutral technologies is not sufficient to meet the needs of a diverse producer constituency. The public sector increasingly should assess the opportunities for R&D and technology transfer for those who are not served by the private sector.
The study committee cites the need for interdisciplinary work that will integrate biophysical sciences, social sciences, and the humanities as an avenue to achieving broad research goals. Farming systems research and extension, which have been primarily implemented in developing countries, illustrate how the integration of social science approaches has broadened the research agenda. In Colombia, collaboration between rural sociologists and bean breeders changed the focus of a bean breeding program from early-maturing to fast-cooking characteristics, based on needs-assessment research conducted with end users (Feldstein and Poats, 1989). The resulting bean varieties were better suited to the needs of small farmers, who were primarily women. In Peru, anthropologists working with agricultural engineers shifted the potato storage research program from complete-dark, off-farm options to partial-light, on-farm storage options. The resulting technologies reduced post-harvest losses and storage costs. More small farmers were able to retain potatoes longer after harvest until prices were high, resulting in increased incomes (Rhoades and Booth, 1982). In the United States context, social science research involving the end users of soil conservation technology also contributed to more effective interactions among strawberry farmers and agricultural engineers in extension and in the Natural Resources Conservation Service (Mountjoy, 2001).
More social science research must be integrated into priority setting. The committee envisions an important role for social science research on agribusiness and entrepreneurial enterprises other than farm management; the costs, benefits, and consequences of technology, including social, human, and community factors; and rural development, including lifestyles and opportunities for individuals and communities. Social science research on farm structure and production systems also can serve as a needs-assessment baseline in research decision making.
It should be noted, however, that many public-sector research institutions lack a significant social science research effort. For example, in 2001, ARS reported only 1 economist, 3 home economists, and no sociologists in its workforce of 1,980 research scientists (USDA, 2001d). Among a total workforce of 4,278 employees in USDA’s Research, Education, and Economics Mission Area, 307 economists, 6 home economists, 29 social scientists, 8 sociologists, and 7 education professionals are reported (USDA, 2001b). Some international agricultural research centers, particularly the International Potato Center, the International Center for Maize and Wheat Improvement, and the International Center for Tropical Agriculture, have successfully and fruitfully involved social scientists in setting research priorities (Ashby and Sperling, 1995). Increased social science research would be a fundamental—perhaps even groundbreaking—public policy response in the United States.
The public sector should use an interdisciplinary approach integrating biophysical, social science, and humanities perspectives to determine structural consequences of research and to assess the research needs of a diverse clientele. The public sector, particularly ARS, should strengthen social science expertise in the areas of setting research priorities and assessing the distributional implications of research and new technology.
The study committee finds that public engagement in research supported with public funds is highly desirable, not only because it will make such work more accountable to the public, but because it has the potential to improve the public sector’s ability to serve a diverse constituency. In the 1940 Yearbook of Agriculture, T.Swann Harding wrote that research conducted inside a professional science culture becomes “celibate” through isolation from the realities of life (Harding, 1940). The effective use of publicly funded research and monitoring of its effects require civic engagement. Research made more accountable to the public extends peer communities, giving scientists the “opportunity to test their work against a wider public and a wider variety of knowledge” (Raffensperger et al., 1999).
Participatory methods developed in the 1980s have been demonstrated to be valuable tools in understanding local people’s needs and priorities with respect to agricultural innovation and technology development (Chambers, 1983; Pretty, 1995). These methods have been used successfully in the CGIAR and in other international contexts to determine the research agenda for plant breeding, crop, and natural resource management (CGIAR, 1999; CGIAR, 2000; WARDA, 2000), and research results from that agenda have benefited small farmers, women, and other underserved groups. Preliminary assessment of watershed management programs indicates that the most successful environmental outcomes have emerged from programs that had facilitated the involvement of local land managers in diagnostic appraisal, planning, implementation, and performance monitoring and evaluation (Thompson and Guijt, 1999). Very little research, however, is available on the structural outcomes of research conducted using participatory methods.
In the United States, a recent survey of agricultural research decision making reveals that stakeholder12 involvement can work effectively (Dyer et al., 1999). Many states have had or are implementing opportunities for stakeholder participation, and federal public-sector agricultural research is starting to engage stakeholders in the agenda-setting process as a result of the legislative mandate in the 1998 Agricultural Research, Extension, and Education Reform Act (U.S.
Congress, 1998). The 22 ARS national programs and CSREES have hosted stakeholders at public workshops for the past 3 years. The National Agricultural Research, Extension, Education, and Economics Advisory Board has engaged stakeholder participation at public hearings around the country.
The committee identified major challenges to participation from conversations with program staff and individuals who had attended its public workshops. There is a lack of coordination in methodology for engaging participation and interpreting stakeholder input in decision making across programs and agencies; it is difficult to obtain representation from regions, sectors, individual farmers, small farmers, commodity producers, and minority institutions; and it is a problem to secure travel funding for individual growers who cannot afford to attend meetings. The committee also found it difficult to obtain information on the Internet about the findings of the stakeholder participation sessions—other than a record of the date, location, and subject area. There was little information to explain how interested parties might become involved in future public workshops. The committee suggests that an examination of public participation strategies could be used to develop more effective approaches at the federal, state, and local levels.
To improve accountability to constituents, the public sector, at both the federal and the state levels, should continue to incorporate the knowledge and needs of stakeholders through genuine public participation in setting priorities for research and in implementing research projects; encourage broad-based participation on research and extension advisory boards to assess the relevance and importance of proposed research and extension programs and to ensure that priority setting is responsive to a variety of needs, particularly those that cannot be met by the private sector; conduct critical analysis and assessment of the methods used for engaging, interpreting, and incorporating stakeholder input into decision making; and take action to make the participation process more understandable and transparent to the public.
Structural Impact Assessments
Data and research on the relationship between public research and structural change are limited. Ex ante impact assessment research on prospective technologic thrusts and ex post research on recently commercialized technologies is most urgent when these technologies are likely to have major impacts on the structure of agriculture, the environment, food safety, or the relations between agriculture and consumers.
Public-sector research institutions, at both the federal and state level, should develop expertise and research programs devoted to analyzing the distributional implications and impacts of agricultural R&D for various groups of producers, using both ex ante and ex post research designs. The study committee endorses the public sector’s earlier efforts in this regard and encourages continued development of this research base.
In this chapter we discussed the structural implications of agricultural research. Empirical evidence suggests that publicly funded agricultural research and development correlates with increases in average farm size, the number of farms, the proportion of large farms as a percentage of all farms, livestock specialization, and off-farm work participation. We also discussed the structural impacts of various types of agricultural innovations and noted the degree of involvement of public research in their development. Using impact on labor and cost per acre as a function of size as criteria for measuring the effects of research on farm size suggests that mechanical innovations and, to a large extent, chemical innovations have more significant effects on size distribution. Innovations in biology are more divisible but can require high fixed costs associated with learning and capital expenses. Managerial innovations appear to have mixed structural impacts. Each category of innovation has varied consequences for structure. It is difficult to measure the exact contribution of public research in innovation because of the close interaction of public and private research and because research often begins in public institutions but is finished and brought to market by the commercial sector.
The discussion of case studies highlighted other issues that link publicly funded research and structural change, including the scale neutrality of research innovations (that is, the ease of adoption of a particular technology, its divisibility, and its potential to benefit large and small producers), the contribution of public research to vertical integration in livestock production, and the contribution of public research to shifts in regional boundaries for modern production agriculture.
A discussion of the structural implications of the research-priority-setting process described three structural implications of using productivity and efficiency criteria for research funding and evaluation. First, in a commodity industry, the benefits of research that increases productivity and efficiency will accrue to lowest cost producers. Second, productivity and efficiency goals fail to consider broad social goals in assessing the benefits of agricultural research. Third, productivity and efficiency criteria fail to adequately assess total resource productivity and efficiency.